Whenever you're ready, walk me through how you'd approach this.
Explain how you would design a system that can handle a 10x spike in traffic during a seasonal event.
Clarify the workload, traffic shape, latency and availability objectives, and whether machine learning inference is on the critical path. Describe the end-to-end architecture, capacity strategy, online and batch components, observability, evaluation, and failure recovery. Address how you would validate the design before the event and safely scale it back afterward.